Customer discovery interviews that reveal real demand

What a customer discovery interview is — and isn't

A customer discovery interview is a short conversation with someone who might have the problem you want to solve. Its only purpose is to learn how they experience that problem today: when it happens, what they do about it, what it costs and what they've already tried.

It is not a sales call, a pitch, a survey or a focus group. You are not trying to convince anyone. If you leave an interview having talked mostly about your idea, it wasn't a discovery interview.

Why interviews come before anything else

Interviews are the cheapest test of the most dangerous assumption most new businesses make: that the problem is real and painful enough for someone to act on it. They cost only time, they can be done this week, and they often change the idea before you've built anything.

They also give you the exact language buyers use. When you later write your offer, the phrases that work best are usually ones interviewees said to you.

Before the interview

Decide what you need to learn

Write one or two learning goals before you contact anyone. For example:

• How often do small shop owners fix catalogue errors, and how long does it take?
• What do they use today, and what do they dislike about it?

If a question doesn't serve a learning goal, cut it.

Choose who to talk to

The right people match your buyer profile and have recently experienced the problem. A useful screening test is a single question: "When did this last happen to you?" If the answer is "never" or "a year ago", they're probably not the right interviewee for this goal.

Aim for 10–15 relevant interviews for a first decision. Stop earlier if answers start repeating; keep going if every interview surprises you.

Find and invite them

Start with your network and communities where your buyers already gather and where such requests are allowed. Keep the invitation short and honest:

Never pretend you're not building something if you are. Saying "I'm researching before I build anything" is honest and lowers pressure.

During the interview

Structure that works

25 questions about real behaviour

Pick the ones that serve your learning goals. You won't use all of them in one interview.

Context

1. Can you walk me through what your role involves in a typical week?
2. Where does [the problem area] fit into that week?
3. Who else is involved when this comes up?

The last time it happened

1. Tell me about the last time you had to deal with [problem].
2. What triggered it that time?
3. What did you do first?
4. How long did it take, roughly?
5. What was the hardest part?
6. What happened as a result — anything that went wrong or cost money?
7. How did you know it was done?

Frequency and cost

1. How often does this come up?
2. Is it getting more or less frequent?
3. What does it cost you when it goes wrong — time, money, customers?
4. Have you ever put a number on it?

Current solutions

1. What do you use today to handle it?
2. What do you like about that? What frustrates you?
3. Have you tried anything else? What happened?
4. Do you pay anyone or anything for this today? How much?
5. If you could change one thing about how this works today, what would it be?

Decision and buying

1. When you last bought something to solve a work problem, how did you decide?
2. Who else would need to agree?
3. What would make you not trust a new solution?

Close

1. Is there anything I should have asked but didn't?
2. Who else should I talk to about this?
3. Would it be all right if I followed up once I've spoken to a few more people?

Questions to avoid

Listening well

• Let silence work. After an answer, wait a few seconds. People often add the most useful detail when the pause gets uncomfortable.
• Follow the emotion. When someone sighs, laughs or says "honestly…", ask "tell me more about that".
• Ask for specifics. "Usually" and "always" hide detail. Ask "what happened the last time?"
• Don't pitch. If they ask what you're building, say you'll share it at the end. Then keep it short.

After the interview

Take notes within an hour

Write the person's own words, with the date, as close to verbatim as you can. Keep your interpretation in a separate column. Our customer interview script has the format.

Know what an interview proves

An interview shows that a problem exists for this person and how they handle it. It does not show that they will pay you. A request for a sample or quote is a stronger signal, and a payment is stronger still. Keep these levels separate in your notes so a friendly conversation is never read as a sale.

Look for patterns, not quotes

After five interviews, list what came up more than once: triggers, workarounds, costs, frustrations. After ten, decide what you've learned against your goals. If the problem rarely happens, costs little or is already well solved, that is a valuable result — it saves you from building the wrong thing.

Turning interviews into a decision

Use the interviews to update the assumptions you started with:

Then write your next test — usually a one-page offer and a sample sent to relevant buyers, with a decision rule. See business idea validation for how to set that rule.

Interviewing business buyers

When the buyer is a business, two extra things matter. First, the person with the problem isn't always the person who pays: an operations assistant may feel the pain while the owner signs off the spend. Try to speak to both, or at least ask who else is involved in the decision. Second, business buyers think in terms of time, risk and cost — ask how long the problem takes each month, what happens when it goes wrong and what they already spend on it.

Remote or in person?

Video calls are fine for most discovery interviews and make it easier to reach the right people. In-person conversations help when the problem is physical — a workshop, a warehouse, a shop floor — because you can see it happening. Either way, ask permission before recording, and if you record, say how long you'll keep it.

Consent and respect

• Say who you are and why you're asking.
• Ask before recording or quoting anyone.
• Don't ask for confidential business data you don't need.
• Keep notes private; anonymise them if you share them with anyone.
• If someone says they'd rather not answer, move on.

People who are treated with respect in an interview are far more likely to become your first customers later.

Common mistakes

• Interviewing only friends. They'll be kind, not representative.
• Pitching in the middle. It turns the interview into a sales call and the answers into politeness.
• Asking about the future. Hypothetical questions get hypothetical answers.
• Not recording originals. Memory edits quotes to fit what you hoped to hear.
• Counting compliments. "That sounds great" is not evidence of demand.
• Stopping at three. Three interviews rarely show a pattern.

How Proofoak helps

In Proofoak, the test package includes a screener, an invitation, a neutral interview guide and a note template linked to your project. Notes go into the evidence ledger with their level — opinion, interest, request, payment — so interviews feed directly into your decision report instead of living in a forgotten document.